Full Article - PDF - International Journal of Conservation Science

Transcrição

INTERNATIONAL JOURNAL
OF
CONSERVATION SCIENCE
ISSN: 2067-533X
Volume 5, Issue 3, July-September 2014: 271-282
www.ijcs.uaic.ro
WERE CASTLE WALLS IN PORTUGAL COVERED WITH
RENDERS? THE CASE OF ARRAIOLOS CASTLE
Pedro GAMEIRO1, Francisco BILOU2, Patricia MOITA3, Cátia MARQUES3,
*
Luís DIAS3, Alexandra FERREIRA3, José MIRÃO3, António CANDEIAS3
1
Matos Gameiro Arquitectos, Av. Elias Garcia 20, 1º dto, Lisbon, Portugal
2
Câmara Municipal de Évora, Praça do Giraldo, 73, Évora, Portugal
3
HERCULES Laboratory, Evora University, Largo Marquês de Marialva 8, Évora, Portugal
Abstract
Arraiolos Castle is one of the medieval fortresses that still stands in Southern Portugal with few
modifications of restoration and preservation interventions. Particularly relevant is the fact that
large areas of the castle walls are covered with mortars and some have gothic inscriptions that
point out to their production in the second half of the 14th century. This work was carried out in
order to verify if the mortars present were the same with each other and with the age of the
mortar inscriptions. The characterization methodology involved a multidisciplinary set of
chemical and microanalytical techniques, as well as radiocarbon dating. The results showed
that all render mortars have similar composition, with aerial calcium magnesium lime as binder
and aggregates with siliceous nature and similar in composition to the rock of the region
around Arraiolos. The radiocarbon blind analysis corroborated the date discovered in one of the
gothic inscriptions found in the renders of the castle walls.
Keywords: Castle walls; Historical mortars; Radiocarbon dating
Introduction
It is common the idea that medieval castles in Portugal always fully revealed the ashlars
or more crudely stacked stone and that these fortifications had the same appearance today as
they had since its foundation. This belief is fuelled either by the romantic view of the castles as
symbol of nationality, in vogue since the nineteenth century, or by the incorrect interpretation of
these defensive buildings, muddying the reading of military heritage. However, disentangling
ourselves from preconceived ideas and based on careful analysis of the most important
medieval documents, in-situ observation of what still remains of the original ancient
fortifications and laboratory analysis of samples of mortars present there - it becomes clear that
castles, for the most part, were covered with renders. Furthermore, maintaining the current
model of interpretation for these buildings constitutes a threat for their preservation [1].
This article relates to the analysis of a case study, in particular – Arraiolos Castle - whose
constitution of stone masonry anticipates the likelihood of being fully rendered.
This fortification arrived fairly ruined to the twentieth century (see Fig. 1a). Although
between 1961 and 1963 the South Tower has been rebuilt, the castle itself was generally spared
*
Corresponding author: [email protected]
P. GAMEIRO et al.
to major restoration interventions, for example, significant cleaning of the coatings was not
made, as happened in many other castles in Portugal.
Fig. 1. Arraiolos Castle: a) general view, b) mortar sampling location,
c) area of castle wall with mortar renders, d) gothic inscription in a mortar
The fortress preserves large areas of renders in all the walls. These renders are possibly,
according to their appearance, to belong to the same historical period. Moreover various gothic
inscriptions (see Fig. 1b) are found especially on the north walls, where the coatings are best
preserved. Although compelling, these would be similar to the findings in other castles, was not
the existence of a particular inscription discovered by Bilou and Branco in 2010 [2] that could
shed some light about the origin of these renders. This inscription, in excellent state of
conservation, registers the making, in 1385, of "twelve niches", a "small restoration in the
walled village” - and reads:
“Era de Mil E IIII(c) E XXIIIº a (-) nos XXIIIIº dias d´aabryl forom fectas Estas XII ameas E
Castelonas G(onçalo) Fernandez pedreyro d´evora Vasco Pirez ho Escriveo”
According to the authors, the modern transcription of the inscription follows:
"Age of thousand and four hundred twenty-three years (year of Christ 1385) twenty-four
days of April, these twelve niches and ‘castelonas’ were made (by) Gonzalo Fernandez, mason
of Evora; Vasco Pirez wrote.”
The above mentioned study focused only on the transcription of this inscription and
others also found in the castle walls and it became clear that all inscriptions were made in the
same period. However, the authors did not extrapolate directly or indirectly, any reflection on
the construction system, nor was the hypothesis of these structures been rendered entirely and
not just in some parts put forward.
In order to prove, definitively, the reliability of the historical descriptions made and to
understand whether all render remains are similar, a multi-analytical study was envisaged, that
included the radiocarbon dating of the mortars from Arraiolos Castle, together with the
chemical and mineralogical composition of a set of samples from different areas of the Castle.
272
INT J CONSERV SCI 5, 3, JUL-SEP 2014: 271-282
WERE CASTLE WALLS IN PORTUGAL COVERED WITH RENDERS? THE CASE OF ARRAIOLOS CASTLE
Experimental
The sampling of mortars is a crucial step that can influence the success of the
characterisation methodology. The selection of the sampling sites was made according to an
architectural systematization study. The location of the samples was mapped onto the
architectural plans and made its photographic record. The size of each sample was the minimum
that could guarantee the success of the analysis. Seven mortars were sampled from different
sites, as shown in Fig. 2 and described in Table 1. All samples correspond to mortar renders
with exception to mortar CA4 which corresponds to a filling mortar.
Fig. 2. Mortar samples collected: a) CA1, b) CA2, c) CA3, d) CA4, e) CA5, f) CA6, g) CA7
http://www.ijcs.uaic.ro
273
P. GAMEIRO et al.
Table 1. Optical microscopy observations.
Sample
CA1
CA2
CA3
CA4
CA5
CA6
CA7
274
Polished surface
Insoluble residue
Observations
Binder colour: light cream colour
Aggregates morphology: sub-angular
and angular
Insoluble residue %: 60
Main
features:
quartz
(hyaline),
feldspars, amphiboles, ceramics
and angular
Main features: quartz, feldspars,
amphiboles, ceramics
and angular
Main features: : quartz, feldspars,
amphiboles
and angular
and angular
and angular
amphiboles
and angular
amphiboles
The mortars were thoroughly examined in the laboratory using a stereo-zoom microscope
and carefully disaggregated to avoid breaking the existing aggregates.
Optical microscopy observations and the corresponding photographic documentation
were obtained by a dark field microscope Leica DM2500 (x100 and x200 magnifications)
equipped with a Leica DFC 290HD camera.
The mortars were thoroughly examined in the laboratory using a stereo-zoom microscope
and carefully disaggregated to avoid breaking the existing aggregates.
Optical microscopy observations and the corresponding photographic documentation
were obtained by a dark field microscope Leica DM2500 (x100 and x200 magnifications)
equipped with a Leica DFC 290HD camera.
Scanning electron microscopy coupled with energy dispersive X-ray spectrometry (SEMEDS) was carried out using a Hitachi S-3700N variable pressure scanning electron microscope
coupled with a Bruker XFlash 5010 SDD Detector. The observations were done under a
pressure of 40 mPa and the current used was 20 kV. For SEM-EDS and optical microscopy the
micro-samples were prepared as polished-sections, mounted in epoxy polymeric resin and
polished with silicon carbide with the granulometry of 1200 µm in order to distinguish between
the different layers and to easily identify the composition of particles by backscattering electron
imaging and also to obtain good quality X-ray spectra and 2D compositional maps.
In order to confirm the mineralogical phases of the mortars, micro-X-Ray diffraction (µXRD) was performed using a BRUKER AXS D8 Discovery diffractometer with a Cu Kα
radiation source and a BRUKER LynxEye energy-dispersive one-dimensional detector. The
diffractogram was obtained in the interval 3-74º 2θ, step of 0.05º, time per step of 1s. The
identification of the phases was performed with the Bruker EVA software package (Version
3.0) using the PDF-ICDD Powder Diffraction Database (International Centre for Diffraction
Data). EVA software enabled also semi-quantitative analysis based on the reference-intensityratio (RIR) method by scaling the maximum intensities of the ICDD PDF patterns to the
observed peaks in the powder pattern.
The mortar’s binder composition and CO2 content was assessed by thermogravimetry
(TGA) and differential thermal analysis (DTA) performed on a Netzsch STA 449F3 Jupiter
analyser, under nitrogen atmosphere, with heating rate of 10ºC/min, from room temperature to
1000ºC. TGA measures the weight (mass) change of a sample as a function of temperature
while the DTA measures the energy changes, represented by endothermic or exothermic peaks
in the DTA curve.
Quantitative analysis is based on the TGA curves (thermograms) while DTA provides
information for the qualitative identification of the components that undergo weight losses [3,
4]. The weight loss of the decomposed main phases was determined by Proteus© software for
the selected temperature ranges.
A fraction of the disaggregated sample was afterwards digested in warm dilute
hydrochloric acid (1:3) to separate the siliceous aggregates from the lime paste. The insoluble
residue was weighed and sieved to determine the particle size distribution of the aggregate
fraction i.e. the siliceous sand.
Taking into account the unique opportunity to study the Arraiolos Castle plasters with
incised Gothic calligraphy, it was considered relevant to proceed with the dating of two samples
275
P. GAMEIRO et al.
using the technique of Accelerator Mass Spectrometry Radiocarbon Dating. The analyses were
performed as outsourcing at the Department of Physics and Astronomy, University of Aarhus in
Finland led by Alf Lindroos and Jan Heinemeier, global experts and forerunners of this
technique for dating mortars [5, 6].
Results and Discussion
Observation by optical microscopy
The first approach to assess the composition of the mortars was the observation with
naked eye of samples and under optical microscope of the polished surfaces and the insoluble
residue obtained after acid attack. Table 1 depicts the observations for each mortar. All samples
have the same type of mortar with a heterogeneous composition and light cream colour that
disaggregates easily, revealing low cohesion. Mortar CA4 exhibits a lighter whitish colour in
respect to the others. Regarding the aggregates, quartz is the dominant mineral. It is also
possible to identify abundant presence of feldspars and plagioclases and some amphiboles. The
composition of the aggregates is characteristic of the rocks of Arraiolos region where
metamorphic rocks of intermediate to high degree abound, like amphibolites, and minor
occurrences of igneous rocks, such as granites. The aggregates have predominantly sub-angular
morphology, being less frequent angular which is characteristic of sands with little transport. It
was also possible to observe the presence of some ceramic fragments.
With the insoluble residue it was also possible to determine the particle size distribution
which is presented in figure 3. The granulometric curves and the insoluble residue amount are
similar to all samples which tends to indicate that all mortars were prepared with the same type
of raw materials (aggregates and binders) and mortar preparation procedures.
Fig. 3. Particle size distributions of Arraiolos Castle mortars (I.R% - insoluble residuum percentage).
276
X-ray diffraction
X-ray diffraction patterns are similar for all the render mortars studied. The results
corroborate that quartz is the main mineral present. These mortars are still characterized by
abundance of feldspars from the group of potassium feldspars and plagioclases and the presence
of considerable amounts of amphibole. Table 2 presents the XRD phase’s semi-quantification
by RIR method.
Table 2. XRD phase quantification.
CA1
CA2
CA3
CA4
CA5
CA6
CA7
calcite
12.3
8.6
5.2
8.1
4.9
3.6
7.8
quartz
42.0
48.2
40.1
33.7
54.6
57.8
51.0
feldspar
23.5
21.0
33.6
32.3
17.0
18.1
17.0
amphiboles
6.2
6.2
2.6
5.4
8.5
6.0
5.2
muscovite
16.0
16.0
18.5
20.5
15.0
14.5
19.0
The results show that the amount of calcite is very low which is consistent with the low
mechanical properties displayed by these mortars. Furthermore, the mortars composition
variability is very low which is consistent with the previous observations that these mortars
were most certainly produced at the same time.
Scanning electron microscopy
Scanning electron microscopy observations allowed a deeper insight on the
microstructure and composition of the mortars. The results show that, for all the render mortars,
the binder morphology is typical of aerial lime mortars (Fig. 4a) and its composition is enriched
in calcium and magnesium (see Fig. 4 b, c). This result is typical of aerial lime mortars of
dolomitic origin. In most cases, calcium and magnesium come apart, indicating the presence of
calcite (CaCO3) and magnesium carbonate compounds formed during the carbonation of a lime
produced from a dolomite [CaMg (CO3)2)] or a limestone with magnesium, according to the
following chemical reactions:
calcination: CaMg(CO3)2 + heat → CaO + MgO + 2CO2↑
slacking:
CaO + MgO + 2H2O → Ca(OH)2 + Mg(OH)2 + heat
carbonation: Ca(OH)2 + Mg(OH)2 + 2CO2 → CaCO3 + MgCO3 + 2H2O
(1)
(2)
(3)
The results also show in some cases an enrichment of magnesium in fractures and at the
binder-aggregates interface (see Fig. 4d). The aggregates are of siliceous origin with prevalence
of quartz and the presence of other aluminossilicates (see Fig. 4e, f). It was also possible to
detect the presence of ceramic fragments.
277
P. GAMEIRO et al.
Fig. 4. SEM-EDS microanalysis: a) lime paste microstructure morphology- sample CA3; EDS 2D elemental
distribution b) sample CA3 – lime paste, c) sample CA5 – lime paste, d) sample CA2 - magnesium enrichment in a
fracture , e) sample CA4 – aggregates composition and calcitic lime paste, f) sample CA5 – aggregates composition
Thermogravimetric analysis
The studied samples display similar thermograms for all the samples besides CA4 (Fig.
5a) mortar for which can be defined similar temperature ranges (Table 3); TG, DTG, DTA
results of Arraiolos castle render mortars (Fig. 5b) display a continuous and smooth weight loss
until 200°C which can be attributed to removal of adsorbed water, followed by the dehydration
of hydromagnesite (3MgCO3.Mg(OH)2.3H2O) between 200-350ºC marked by peaks around
250ºC. A mass loss at around 396-429ºC is attributed to the dehydroxilation of hydromagnesite.
Calcium carbonate is marked by a pronounced weight decay between 711 to 773 ºC. The drops
between 450-600ºC for CA5, CA6 and CA7 samples can be attributed to decarbonation of
278
hydromagnesite. These results corroborate the data prior observed by SEM-EDS that showed
the presence of magnesium and calcium in the binder.
a
b
Fig. 5. Thermograms (TG, DTG and DTA curves) from Arraiolos Castle mortars; a) calcitic filling mortar (sample
CA4); b) representative thermogram from magnesium render mortars (CA6 sample).
Contrary to the other thermograms CA4 doesn’t show mass losses that can be attributed
to the presence of magnesium carbonate. The only significant drop is attributed to
decarbonation of calcite with a peak at 763ºC (Table 3). This is consistent with the fact that
CA4 is a filling mortar while the others correspond to render mortars.
279
P. GAMEIRO et al.
Table 3. Weight losses (%) as a function of temperature ranges (ºC) for render mortars.
ref.
CA1
CA2
CA3
CA4
CA5
CA6
CA7
weight losses (%) as a function of temperature ranges (oC)
40-200
200-350
350-600
600-1000
1,30
1,87
5,63
10,78
1,23
1,67
6,00
7,39
1,02
1,15
3,30
6,55
1,82
9,73
40-200
200-350
350-455(50)
455(50)-600
600-1000
1,17
1,46
2,21
2,68
5,50
1,16
1,30
1,68
4,02
5,63
1,38
1,18
3,63
4,05
4,98
Radiocarbon dating
One has to take into consideration that the dating of mortars by the radiocarbon method
is very recent and requires special conditions such as the absence of calcite aggregates, as these
necessarily lead to obtaining a production date prior to the date of the mortar, and the absence
of dissolution processes and re-carbonation because this results in the incorporation of carbon
dioxide in the structure and consequently at a later date to the mortar. The composition of the
Arraiolos mortars which had no evidence of calcite aggregates or dissolution processes and recarbonation, and the fact that it is a technique used successfully in several case studies
prompted the team to consider its application and therefore, three samples were selected for
radiocarbon dating using AMS. The dating was only possible in two samples.
Radiocarbon ages obtained are reported in conventional radiocarbon years BP (before
present = 1950) according to the international convention [7]. Thus, all calculated radiocarbon
ages were corrected. Ages in calibrated calendar years were obtained from the calibration
curves of Reimer et al. [8], using Oxcal [9] v4.1 program calibration curve using the terrestrial
calibration IntCal09. The probability method was used to calculate the calibrated age ranges
corresponding to 68.2% probability (1x standard deviation) and 95.4% probability (2x standard
deviation). The radiocarbon dating curve is presented in figure 6 and the results in table 4.
Table 4. Radiocarbon calibrated dates.
sample
1
callibration curve and
correction
IntCal09 (Atmospheric)
2
IntCal09 (Atmospheric)
Calibrated data
68.2% probability
1324-1346 (31.4%)
1393-1416 (36.8%)
1423-1445 (68.2%)
Calibrated data
95.4% probability
1308-1362 (47.9%)
1386-1430 (47.5%)
1408-1456 (95.4%)
The calibrated dates obtained by the radiocarbon method demonstrate that the existing
mortar coating on the renders of the Arraiolos Castle walls are from the late fourteenth / early
fifteenth century. These results are consistent with the gothic inscription reporting the date of
1385, and a clear evidence of the use of mortar coating in the castle walls in this period.
Furthermore, and taking into consideration the inscription date, it is also a significant result on
the usefulness of radiocarbon dating of mortars.
280
Fig. 6. Radiocarbon determination as a function of calibrated data
Conclusions
The multianalytical methodology applied showed that Arraiolos Castle mortars have
similar compositions with magnesium-calcitic binder, and siliceous sands with compositions
similar to the rocks from Arraiolos surroundings. It was finally possible to understand that
renders in areas located far from each other are similar and originate from the same period which would lead to conclude that most probably, the whole castle walls had renders that
covered the whole surface. Particularly relevant was the fact that radiocarbon dating is
consistent with the date discovered in an incisive inscription found in one of the mortars of the
castle which points out its production in the second half of the fourteenth century beginning of
the fifteenth century.
Acknowledgements
The authors which to acknowledge QREN/PORA Regional Framework Programme for
financial support through project IMAGOS
References
[1] P. Gameiro, O Semblante Original das Fortalezas Medievais em Portugal, PhD Thesis,
University of Granada, delivered in April 2014, awaiting defence.
[2] M. Branco, F. Bilou, Inscrição e Grafitos Medievais no Castelo de Arraiolos, Arraiolos:
Câmara Municipal de Arraiolos, 2011.
281
P. GAMEIRO et al.
[3] J. Hughes, K. Callebaut, Historic Mortars: characteristics and tests, Practical sampling of
historical mortars, Proceedings of the RILEM International Workshop, Paisley, 1999,
pp.17-26.
[4] M.R. Veiga, J. Aguiar, A.S. Silva, F. Carvalho, Methodologies for characterisation and
repair of mortars of ancient buildings, Proceedings of the 3rd International Seminar
Historical Constructions, Guimarães, 2001, pp. 353-362.
[5] J. Heinemeier, H. Jungner, A. Lindroos, A. Ringbom, T. von Konow, N. Rud, AMS 14C
dating of lime mortar, Nuclear Instruments and Methods in Physics Research B, 123,
1997, pp. 487-495
[6] J. Heinemeier, Å. Ringbom, A. Lindroos, Á.E. Sveinbjörnsdóttir, Successful AMS 14C
Dating of Non-Hydraulic Lime Mortars from the Medieval Churches of the Åland Islands,
Finland, Radiocarbon, 52(1), 2010, pp. 71-204.
[7] M. Stuiver, H.A. Polach, Discussion of reporting 14C date, Radiocarbon, 19(3), 1977, p.
355.
[8] P.J. Reimer, M.G.L. Baillie, E. Bard, A. Bayliss, J.W. Beck, P.G. Blackwell, C. Bronk
Ramsey, C.E. Buck, G.S. Burr, R.L. Edwards, M. Friedrich, P.M. Grootes, T.P.
Guilderson, I. Hajdas, T.J. Heaton, A.G. Hogg, K.A. Hughen, K.F. Kaiser, B. Kromer,
F.G. McCormac, S.W. Manning, R.W. Reimer, D.A. Richards, J.R. Southon, S. Talamo,
C.S.M. Turney, J. van der Plicht, C.E. Weyhenmeyer, Intcal09 and marine09 radiocarbon
age calibration curves, 0–50,000 years CAL BP, Radiocarbon, 51(4), 2009, pp. 11111150.
[9] C.B. Ramsey, Bayesian Analysis of Radiocarbon Dates, Radiocarbon, 51(1) 2009, pp.
337-360.
Received: January, 30, 2014
Accepted: August, 14, 2014
282

Full Article - PDF - International Journal of Conservation Science

Transcrição

Documentos relacionados

PR 2 LSA - Câmara Municipal da Lousã

Read Full Paper

Mortar and Pestle Set

1. Description: Biomassa is an innovative high performance mortar

Job Specification – Corporate Responsibility Executive

Non-literal Language Use and Coordination in Dialogue

h[$torioatog - The Henry County Historical Society and Museum

Print this article - Journals at the University of Arizona

Crossing the Luzon Strait - National Museum of Prehistory

Programme FMD 2012 - Materiais Diversos

Apresentação 1 - Sociedade Portuguesa de Psicologia da Saúde